Evaluation of Brain MRI Alignment with the Robust Hausdorff Distance Measures (original) (raw)
Related papers
Robust registration of volumetric image data
Head motion during fMRI experiments continues to be a significant problem for analysis, producing artifacts that can severely degrade image quality and make interpretation difficult and inaccurate. Such movement artifacts can lead to reduced statistical significance when detecting true activation or may lead to the creation of spurious activations. To address the problem of interscan motion, a retrospective image registration method has been developed. The registration technique is based on the use of non-linear deformation fields, the least-trimmed-squares robust estimator and Procrustes analysis. The registration algorithm was validated using simulated anatomical MRI volumes and real fMRI datasets. The registration technique is robust in the presence of large amounts of noise; and the experiments show that the method gives accurate estimations of motion up to 5 mm translation in all three directions and 5 degrees rotation around the three axes. The correction procedure also yields...
Comparison of Point Similarity Measures for Atlas-based Registration of MRI Brain Images
Conference proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference, 2005
High-dimensional deformable registration of MRI brain images is presented here. The deformation is driven by local forces estimated from point similarities based on joint histogram and with the use of prior information obtained from tissue probability maps available in selected commonly used brain atlases. Three point similarity measures are tested in an experiment with data obtained from standard Simulated Brain Database.
IRJET, 2022
Image registration is done prior to image fusion. Medical and biomedical images are taken as an example for CT images which rely on reference image. The estimated time and error rates are calculated based on the transformation of images. This paper introduces the calculation of performance quality metrics based on the pictures taken during the MRI and CT of a brain. The approach is to create a method for stiff CT and MRI image registration using wavelet image fusion.